MQA more answers, maybe too many for some people.

[Blumlein 88]

http://www.stereophile.com/content/mqa-questions-and-answers#TjIfWkFPCfBITGf9.97

More answers than you can shake a stick at.

 

[ceedee]

Wow! I've never seen an advertisement that needed its own glossary before!

 

[Sal1950]

Bob looks impressive in hip "jeans w/sport coat and open shirt" style but I think the white lab coat bit would have worked better. He is stylin in with the Trump bad comb-over hairdo
How come he's giving us the "whatsamatteryu" Italian hand gesture?
That was one hell of a detailed write-up. But it needs the glossary and more ceedee!
Not one in ten thousand subscribers has any idea what he's talking about, including me. Could all be pure technobabble BS as far as they know. LOL

 

[Liberal Democracy]

Arrr shutupa ya face

 

[ceedee]

Not one in ten thousand subscribers has any idea what he's talking about, including me. Could all be pure technobabble BS as far as they know. LOL

That's what I was thinking. How many readers of Stereophile are going to understand one tenth of that? It seems like he is trying to overwhelm us with large words and prodigious technical references. Perhaps a good audio demonstration would be better? Let's see what's actually audible.

But "many recording and mastering engineers have testified!"

Many recording and mastering engineers have testified that MQA improves very considerably on the conventional methods, recreating the sound they actually hear or remember from the original session or, in the case of archive material, the sound from an analogue tape recorder.

 

[amirm]

Bob Stuart is a friend and professional colleague. He is one of the few in the high-end with proper background and education in signal processing and psychoacoustics. I read through the article and it is superb. Normal papers he writes as he mentions are for his peers at AES and hence assume similar background to him. In these articles he is trying to bring that one level down. As said, that is hugely insufficient to get the normal audiophile to grasp any of it. But for me, it does fill the gaps and provides excellent references and tutorials on core signal processing topics. So with my biases about him included, this is a superb effort .

 

[Cosmik]

So on the one hand we have a consensus that "phase doesn't matter" in speakers (and timing) - and we are talking about millisecond levels there - and on the other we have a load of words that say that microsecond accuracy is essential. I disagree with both!

 

[ceedee]

I read through the article and it is superb.

I can't say I understood everything in the article, but I trust your opinion on it. If his goal was to dumb it down, he failed miserably at that. I would bet most Stereophile readers don't even understand Nyquist!

However, it still sounds like a solution in search of a problem. Proper blind testing could demonstrate otherwise, though.

I'd definitely be interested in an increase in fidelity if possible, but let's hope it's not another format that will require a meta-analysis in 15 years to 'prove' that it's audible to some people.

 

[amirm]

So on the one hand we have a consensus that "phase doesn't matter" in speakers (and timing) - and we are talking about millisecond levels there - and on the other we have a load of words that say that microsecond accuracy is essential. I disagree with both!

That's not what the article says. Indeed it clarifies that MQA is NOT about absolute timing accuracy at all. It is about spreading in time the power of an impulse, not that the impulse itself is moved.

I also need to clarify that the value I see in the article is in the tutorial aspect of signal processing and psychoacoustics. I am not expressing a point of view about merits of MQA.

 

[amirm]

I can't say I understood everything in the article, but I trust your opinion on it. If his goal was to dumb it down, he failed miserably at that. I would bet most Stereophile readers don't even understand Nyquist!

However, it still sounds like a solution in search of a problem. Proper blind testing could demonstrate otherwise, though.

I'd definitely be interested in an increase in fidelity if possible, but let's hope it's not another format that will require a meta-analysis in 15 years to 'prove' that it's audible to some people.

What he says in theory makes some sense. But as you, I don't hold any hope that a) audiophiles actually hear these deficiencies and b) the market will allow a new format to be created. The typical customer for MQA is turned off by the "lossy" coding of higher end of the bandwidth and with them gone, there is nobody left to buy into it.

 

[Sal1950]

Admitting I'm dumb as a box of rocks here the question still boils down to, is there any audible gain in SQ available? Or is MQA a cool technical exercise for the engineers that could also offer a path to additional profits from the sale of new gear and encoded music? It's obviously a very clever compression scheme but what really are it's over all effects on SQ? We don't really need a better compression code today and it becomes less valuable by the day.
I'm still very confused. LOL

 

[Cosmik]

That's not what the article says. Indeed it clarifies that MQA is NOT about absolute timing accuracy at all. It is about spreading in time the power of an impulse, not that the impulse itself is moved.

Spread? Absolute timing accuracy? I see the two things as part of the same phenomenon. The typical passive speaker is a superb impulse spreader, and in terms of absolute accuracy it's anyone's guess where the impulse actually lies in amongst the phase shifts and timing errors. A time aligned, phase corrected speaker may be somewhat better - but we are told that this is wasted effort because an audiophile consensus says it's inaudible.

And what happened to the audiophile consensus that in real rooms, the reflections render all talk of phase and timing (= impulse spreading) irrelevant? - not something I ever agreed with, but ironic in light of the fuss about MQA.

I am just enjoying the irony.

 

[amirm]

Spread? Absolute timing accuracy? I see the two things as part of the same phenomenon.

They are not. I can have an impulse happen at 1.0001 seconds or 1.0002 seconds. That would be a timing shift and accuracy error assuming one or the other is the right one.

Energy smearing would occur if instead of silence prior to that impulse at that time, you took some of the impulse power and spread it before and after the pulse itself. The pulse will remain at its precise time so timing accuracy is maintained. But energy distribution is not. Audible effect is not at all like timing variation.

 

[Blumlein 88]

They are not. I can have an impulse happen at 1.0001 seconds or 1.0002 seconds. That would be a timing shift and accuracy error assuming one or the other is the right one.

Energy smearing would occur if instead of silence prior to that impulse at that time, you took some of the impulse power and spread it before and after the pulse itself. The pulse will remain at its precise time so timing accuracy is maintained. But energy distribution is not. Audible effect is not at all like timing variation.

So lets say we go with 96 khz sample rates and the related ringing-like behaviour is all going on at 48 khz which your ear will not begin to respond to. Only the wider impulse will cause your ear drum to move an appreciable amount. It is true you have moved some of the energy that would have been in that impulse reducing its peak, but everything about filters and limited FR of hearing makes me think the ear drum would have also lost about the same amount of energy in the impulse were it presented with the real thing. That amount of energy is at a frequency it can't respond to in the first place. The band limiting has preserved the part the ear can respond to if the band is wide enough.

 

[Cosmik]

They are not. I can have an impulse happen at 1.0001 seconds or 1.0002 seconds. That would be a timing shift and accuracy error assuming one or the other is the right one.

Energy smearing would occur if instead of silence prior to that impulse at that time, you took some of the impulse power and spread it before and after the pulse itself. The pulse will remain at its precise time so timing accuracy is maintained. But energy distribution is not. Audible effect is not at all like timing variation.

As I said, if your speaker is spreading the signal in a frequency-dependent way (with two or three drivers all of which have different delays), where is the centre of any 'impulse'? It has been smeared so much that no one can say where its centre is any more. The timing inaccuracy and energy smearing are off the scale compared to anything that MQA is claiming to fix.

 

[fas42]

I can't say I understood everything in the article, but I trust your opinion on it. If his goal was to dumb it down, he failed miserably at that. I would bet most Stereophile readers don't even understand Nyquist!

However, it still sounds like a solution in search of a problem. Proper blind testing could demonstrate otherwise, though.

Have to agree here - I used to digest this sort of techno talk with ease in the ol' days - now, my head just starts exploding the further in I go! The examples at the 2L website say there's monkey business going on - looking at what's in the waveforms of the versions of a particular piece tells me that a fair bit of fiddling is happening, which is going to be audible, regardless - so, which is correct?

 

[RayDunzl]

Benchmark Media's John Siau weighs in on MQA

IS MQA DEAD ON ARRIVAL?

https://benchmarkmedia.com/blogs/application_notes/163302855-is-mqa-doa



I guess they won't be sending me a firmware upgrade any time soon.

 

[amirm]

Benchmark Media's John Siau weighs in on MQA

IS MQA DEAD ON ARRIVAL?

https://benchmarkmedia.com/blogs/application_notes/163302855-is-mqa-doa
I guess they won't be sending me a firmware upgrade any time soon.

Very good (and damning) write up on MQA!

 

[DonH56]

Finally made time to read through the original article. Quite the dissertation. Very interesting approach and IMO very hard to explain without delving rather deeply into technical details that most audiophiles (and reviewers) will fail to follow. What I did not have time to do was to try to find the references explaining the need for much greater time resolution than in current schemes, and that seems to be the primary basis for MQA. It's an interesting argument but one I am not competent to debate. A few years ago I was able to calculate the difference in time we (humans) can distinguish and it was much, much lower than I would have guessed but IIRC still well above the aperture time of a 16-bit converter. And there is always the debate about how much it matters in the real world, with real musical signals and such. The fact that a cymbal crash contains signal content above 20 kHz does not mean that I can hear it, which is acknowledged in the paper, but how the lack of time resolution degrades the sound is unclear to me. I do not know how important the temporal smearing described in the paper is in the real world, i.e. how well I can hear it. Be interesting to test with and without MQA to see, but perhaps my ears aren't up to the task. The claim is most anybody can hear the differences so I wonder (do not know) how much of the debate is based upon "blind" rejection without actually listening tests.

As for MQA being DOA, I suppose it depends upon how well they can market it. I am a little surprised at the apparent level of resistance, but suspect a lot of that is from a combination of misunderstanding of the technology and inbred rejection of anything that appears to make assumptions about what we can and cannot hear. The latter I base upon the high level of concern related to the use of bits below the noise floor. Nobody wants to be told something is below their noise floor and people continue to reject studies showing e.g. two cables can sound the same; "not in my system!" That said, I do not claim to have the hearing many reviewers and others claim to have achieved, it is quite clear others hear what I cannot, and it is impossible to resolve a debate that takes on the nature of a religious war wherein beliefs and science collide.

IMO - Don

 

[RayDunzl]

As for MQA being DOA, I suppose it depends upon how well they can market it.

End User Marketing:

I'm ready! Let's play! What do I need?

A disc, wait, it doesn't come on disc so I have to have streaming, and it has to be streaming without any modification...

And nobody is streaming it, are they?

And I have to have a device to decode it because I don't...

And I have to abandon my room correction since it sits in the digital path and modifies the bits and bytes...

And I can't use my digital volume control since it mangles digits, so I'll have to make sure I have that worked out...

And if it is "high frequency" stuff that makes it really good I can't hear that any way you serve it...

That feels like enough to dissuade me at the moment, since I'm fat, dumb, and happy with what I have going now.

---

Maybe Meridian just has to market it to the vendors:

You'll need a license, and some new hardware, and maybe some more licenses, and probably some software, with licenses and non-disclosure agreements, and that's it! You're in business! Did we mention the licenses? We don't seem to have your NDA on file.

Except not. See above.

---

Although I appreciate "good sound" it is "good music" that I primarily chase, and I have that in spades now. Are they going to remaster all these obscure titles I have?

I know, I don't count, so, we'll see where it goes.

---

My Bad Idea for the day:

How about distributing some gussied up (excuse me: Master Quality Authenticated) 96/24 files we can actually play, and maybe get a taste of what they're up to?